This application addresses broad Challenge Area (03) Biomarker Discovery and Validation, and specific Challenge Topic 03-CA-101, Fingerprints for the Early Detection and Treatment of Cancer. One of the most fundamental hurdles for the effective treatment of cancer is the heterogeneity of the disease. Because of differences in the molecular details responsible for the development and progression of a particular tumor, there can be profound differences in the course of disease even for tumors that appear similar or identical by standard pathological analysis. This problem is becoming more acute as new therapeutic compounds are developed that target specific signaling pathways or molecules. Such targeted therapies may be highly effective for only a small percentage of tumors. Reliable methods to predict which tumors will respond are thus essential if such therapies are to be used effectively. New molecular diagnostic methods to classify tumors, and thus predict response to specific therapies and provide prognostic information on the likelihood that the disease will spread or recur after therapy, hold great promise for more effective cancer therapy. Such methods, if available, will allow the physician and patient to choose the most effective course of treatment, while avoiding ineffective or unnecessary treatments that diminish quality of life for patients and financially burden the healthcare system. Many key biological activities of tumors are controlled by tyrosine phosphorylation, and the dysregulated activation of tyrosine kinases is well known to underlie the development of many tumors. Thus profiling the global state of tyrosine phosphorylation of a tumor is likely to provide a wealth of information that can be use to predict the behavior of the tumor. However current methods to analyze tyrosine phosphorylation are not amenable to the comprehensive analysis of large numbers of human cancer specimens. In this proposal, we will use a novel phosphoproteomics platform, SH2 profiling, to profile non-small cell lung carcinoma (NSCLC) samples from human patients. NSCLC is a devastating disease that kills over 160,000 people per year in the U.S. Certain tyrosine kinases are frequently activated in NSCLC, and new drugs have been developed that target these kinases. In this project we will test whether SH2 profiling can be used to classify NSCLC for prediction and prognosis. If successful, these studies will set the stage for development of clinical tests that can be used to guide more effective treatment for lung cancer and other tumors. PUBLIC HEALTH RELEVANCE: Tumors vary greatly in the course of disease and in how they respond to specific therapies. Molecular diagnostic methods that can be used to classify tumors and predict their behavior have enormous potential to both increase the effectiveness of treatment, and decrease the suffering and cost associated with unnecessary or ineffective treatments. In this proposal, we will us a novel method to profile the global state of tyrosine phosphorylation in lung cancers, and assess whether it provides useful information that could be used to guide treatment.